Is diversification for idiots?

bunny

Concentration is cool

Value investors are prone to make outsized bets on single stocks or have extremely concentrated portfolios.

Mark Cuban has a more blunt view of diversification:

“Diversification is for idiots.”

Warren Buffett, the master himself, made many concentrated bets throughout his investing career. The most famous example is American Express during his partnership days. After the salad oil scandal in the 1960s, American Express stock plunged. Most investors thought the salad oil scandal would put American Express out of business and the stock fell.

Buffett realized that American Express would survive and knew it was a fantastic franchise. He moved an astonishing 40% of the partnership’s assets into American Express, sending the partnership’s returns from great to extraordinary. Buffett would repeatedly make concentrated bets throughout his career when he had conviction. He took on significant, concentrated positions in companies like GEICO and Coca Cola.

Buffett had this to say about diversification:

“Diversification is protection against ignorance. It makes little sense if you know what you are doing.”

Munger has this to say:

“The whole idea of diversification when you’re looking for excellence is totally ridiculous. It doesn’t work. It gives you an impossible task.”

Because value investing has come to mean whatever Warren Buffett and Charlie Munger say it is supposed to mean, most value investors take their advice to heart and make big, courageous, outsize bets on their “best” ideas.

The Other Side of Concentration

Concentration is cool until it is not. Legends like Bill Ackman built up early, stunning returns by concentrating on their absolute best ideas. This worked well until Ackman’s best ideas were Valeant and shorting Herbalife.

Concentration worked out well for Joel Greenblatt, whose early hedge fund used extremely concentrated portfolios of spin offs and special situations to deliver stunning returns.

Ben Graham recommended holding 30 stocks and believed this would provide adequate diversification. The eggheads in academia actually agree with him on this point, and most the research shows that an ideal portfolio is around 20-30 stocks.

Other successful value investors who are more in the Graham state of mind also have diversified portfolios. Walter Schloss owned over 100 stocks. Seth Klarman currently has 35 holdings. Irving Kahn also stuck to Graham’s advice and owned 20-30 positions at any given time.

The Evidence

For me, I’m more interested in the evidence of what works so I conducted backtests on a value portfolio in an effort to identify the ideal number of holdings. I backtested the returns and characteristics on a simple EV/EBIT portfolio. I began with the insane portfolio of buying the single cheapest stock in the S&P 1500. I then added the next cheapest, creating portfolios ranging from 1 stock to 30 to determine the ideal size.

This is pulled from an S&P 1500 universe, the backtests are conducted since 2005, and the portfolios are rebalanced annually.

Returns

The return characteristics of different portfolio sizes are all over the place, but they all beat the market over the long run. Strangely, they start out high, bottom out around a dozen stocks, and then start to improve.

portfoliosize

Volatility

I don’t think volatility is risk. I believe risk is the probability of losing money permanently. However, I also believe that volatility accurately measures how big a container of Tums that you should have at your desk.

I also don’t think a Sharpe ratio is the accurate representation of how decent a portfolio is.

With that said, here are the results:

standarddeviation

sharpe

In terms of volatility, most of the benefits come from the first dozen positions. However, in the context of EV/EBIT, the Sharpe ratio tends to max out around 25 positions.

Maximum Drawdown

My view of risk is less about volatility and more preventing blowups that will permanently impair my capital.

The bigger the blowup, the harder it will be to recover from it. In mathematical terms, large drawdowns are increasingly more difficult to recover from.

drawdowns

The deeper the drawdown, the bigger that the bounceback needs to be. Once the drawdown exceeds 50%, it becomes nearly impossible to recover from. Drawdowns of the 80% magnitude are almost certainly fatal to investment results. Common sense tells us that concentration will lead to years where returns are remarkable and that this will likely be balanced out by deep and painful drawdowns later on. The evidence backs this up.

maxdrawdowns

Concentrated portfolios are prone to some genuinely epic drawdowns. For the insane 1-stock portfolio, the maximum drawdown was 90%. Over the long run, the 1-stock cheap portfolio delivers a return similar to all of the others, but you will likely jump out of a window or develop a drug problem before you can actually enjoy those returns.

It appears that the first 15 positions can at least get the portfolio down to the market’s maximum drawdown.

Summary

  • Most EV/EBIT portfolio sizes beat the market, from the ultra-concentrated to 30 or more stocks.
  • In terms of reducing volatility, the first dozen stocks provide most of the benefits of diversification.
  • It takes at least 15 stocks to reduce the odds of a portfolio blowup that will exceed the maximum drawdown of the S&P 500.
  • Sharpe ratios tend to be maximized in a value portfolio around 25 positions.
  • Diversification is not for idiots.
  • Ben Graham was right (as usual), and the ideal portfolio size is somewhere around 20-30 positions.

Random

“Killing Eve” is a great show.

PLEASE NOTE: The information provided on this site is not financial advice and it is for informational and discussion purposes only. Do your own homework. Full disclosure: my current holdings.  Read the full disclaimer.